Thermaltake TR2 RX 750W - Fixed +5v Stby, now no +12V and +5V is low

[rodd]

Dear all,
I am trying to repair the ATX power supply.
First I fixed the +5V stby, which was zero due to a shorted TNY279 and an open 2R2 resistor.

Just in case I replaced the PWM control IC (PS223) and the PFC control IC (CM6800).

Now the supply turns on, the +3.3V is fine, but the +5V measures only 2.5V and there is no +12V.

I have checked every diode I could find with a DMM (in circuit) and all the capacitors (in circuit) no shorts, low ESR.
In despair I removed the low voltage diodes which were all mounted on the same heatsink to test the out of circuit; all teste GOOD.
I used a bench power supply adjusted to the value of each voltage rail and connected it to the PS outputs- the current "drain" was low in every case, so no shorts.

I ran out of ideas. Any suggestion?
Thanks

[Swake]

Transformers or coils overheated? Might have shorted some windings

[MathWizard]

Do you have an oscilloscope ? Do the PFC or PWM gate drives from the CM6800 start up at all ? Some DMM's might have good enough frequency measurement, and would see ~5V average, but at 50kHz or a bit more.

The PS229 monitors the output currents and voltages, and is tied into the motherboard on/off circuit. The CM6800 is for PFC stage and main PWM mosfets. The TNY chip is for making 5Vstby and ~12V aux for all the control circuits like opto's, on the primary side, all that's ultimately powered right off the main bulk capacitor.

What are the opto-couplers doing, they are the main external parts that would tell the chip to shutdown if there's a problem. Both controller chips sense a few things that can make them shut themselves down.

Is the CM6800 actually getting Vcc ? If 5Vstby is there, then there should be roughly 12V auxiliary, and there's a little circuit to switch on or off the CM6800, including if there's a problem from the secondary side.

What kind of rectifier is on the secondary side ? Is it diodes, or MOSFET's in a synchronous rectifier ? And where does the 5V and 3.3V output rails come from, the rectifier, or from another step-down converter, usually on another little PCB, fed with 12V. They could have a problem too.

[rodd]

Yes I have a scope.
I will check the waveforms tomorrow, since today is father's day here in Brazil.
Please take a look at the pictures of the PCB (top and bottom).
The heatsink with the secondary rectifiers was removed so I could check the diodes out of circuit.

[rodd]

top view

[MathWizard]

Yeah so it's diode rectification, and no 12V to 5/3.3V buck converters.

I'd see what the output rails get to on start-up, set your trigger for rising to a volt or two and use single capture if it's a digital scope.

If the opto-'s are working correctly, you might be able to see if one of them changes right before it all shuts down. There's 1 for the 5Vstby and AUX rail, usually the PWR_ON is signalled to the primary side by that. Then usually 1 for if the main outputs have a problem, and another 1 for actually adjusting the outputs as the load changes, sometimes there's a 4th 1. But most the circuits I've seen for these, just act like comparators, and output a high or low signal to the other side.

[rodd]

Thank you for your reply.
Yesterday I realized that I have installed the PS223 backwards in the socket.
When I installed it right there is no voltages at all, so I think it is fried.
I am waiting the new replacement part to arrive to continue the repair.

[MathWizard]

That's not good. I've burned some of a controller PCB like that 1, with a hot air gun. And I have another controller PCB I took all the caps off, years ago. I'll have to study a bit more before trying to calculate values for replacement caps.

Here a schematic in LTSpice that might help, from some 750W Thermaltake I almost fixed. There's some SMD cap's I don't trust anymore, I probably cooked them too long trying to desolder other stuff.

[rodd]

Thank you for your reply and the info MathWizard,
I have bad news.
After replacing the PS223 IC the Power Supply does not turn on anymore and the voltage across the PFC capacitor dropped to 173V. Is the CM6800 fried again?
I have some basic questions:
1 - In a previous post it was said that both the PS223 and the CM6801 are powered by a 12V supply. Is this volatge always "ON" as the +5V Stby?
2 - in case the 12V is present, the PWM signal at pins 11 and 12 of the CM6800 are always present?
3 - I am confused about how these two ICs work together, specially if PS223 detects a fault. would there be a PWM signal at pin 11 and or 12 of CM6800?
4 - Is there a way to test the basic function of these two IC out of circuit so I can be sure that there is a fault elsewhere?

Thanks

[rodd]

Hi,
More news from the "front".
I inslalled the PS223 backwards again and the PS starts and all the voltages are ok.
This means that the CM6800 and all "power semiconductors" are ok.
So I guess that the problem is located on the small pcb where the PS223 is installed.
As I asked in my previous post, I do not understand very well how the PS223 works but it seems that installing it bacwards causes the ok signal to be present.
I have already checked every resistor (on board) and the readings seem ok.
The there are two to92 devices and one diode on the component side.
unfortunately the markings of one of the devices is covered by glue, so I can't tell if it is a transistor or a 3 pin reference.
There is also a smd zener diode which I don't know if it is working but it tested good as a diode.
Is there a schematics available for this board?
Any idea of what might be triggering on the FAULT status if it seems that there is no short at the outputs?
Thanks.

[rodd]

Thank you all for the replies.
I could fix it!
There were two problems after I fixed the 5V stby rail.
1 - I damaged the PS223 by installing it backawards
2 - I broke my thermaltake power supply tester during the tests, so the start button in it won't start the PS.
This one was very difficult to pinpoint since I thought that the fault was in the power supply, not in the tester

[MathWizard]

Ok you got it fixed, yeah most every pin of the PS223 is behind some resistor and/or capacitor network, and the max voltage of anything on the secondary side is maybe 13V at the transformer output before the rectifier diodes. So it's probably hard to damage the chip, most pins anyways.

The 12V aux voltage, it comes from the same transformer that makes the 5Vstby. That's usually run by a little 8-pin switcher chip. If the supply is plugged in, and the switch on the back of the PSU is on, and there's no problems, then the little 8-pin switcher should always be running. But that 12Vaux voltage is kept away from the CM6800, with a BJT or 2, and an opto-coupler, until the PSU-on button on the PC is pressed. That tells the PS223 or some little circuit around it, to open or close the opto-coupler, and some BJT, which then allows 12Vaux to reach the CM6800 Vcc pin.

If the CM6800 doesn't have a problem, I think both gate drive pins should be running. Sometimes under low loads, like no computer hooked to the PSU, it might not run the boost drive?? IDK, I should learn more about these chips too. A lot of these switching regulator chips, and controllers, will have low power /idle modes, where they run at a lower frequency, or burst-mode, where they run for a little bit, then wait for a little bit, and repeat.

The PS223 or similar chip, usually can run from 5Vstby, but uses the main 12V output rail for power, once that's up and running.

[rodd]

Thank you for your reply, MathWizard.
I have sent you a PM asking for the schematichs in another format.
Have you seen it?